k-omega and k-epsilon turbulence model. Why different?

 Register Blogs Members List Search Today's Posts Mark Forums Read

 April 20, 2011, 02:24 k-omega and k-epsilon turbulence model. Why different? #1 Member   Casey Join Date: Jun 2009 Posts: 98 Rep Power: 17 Hi all, I am confused about the difference between the k-epsilon and k-omega turbulence models. If [math]\omega[\math] = k/, then how are the two equations so different? Can someone explain the models for me term by term and show why the k-[math]\omega[\math] model is better suited for compressible flows and modeling shear flow spreading. Why is the k-[math]\omega[\math] model less sensitive to rapid changes in mean shear rates? Thanks, Casey Far likes this.

April 20, 2011, 03:44
#2
Senior Member

Raashid Baig
Join Date: Mar 2010
Location: Bangalore, India
Posts: 135
Rep Power: 16
Quote:
 Originally Posted by meangreen Hi all, I am confused about the difference between the k-epsilon and k-omega turbulence models. If [math]\omega[\math] = k/, then how are the two equations so different? Can someone explain the models for me term by term and show why the k-[math]\omega[\math] model is better suited for compressible flows and modeling shear flow spreading. Why is the k-[math]\omega[\math] model less sensitive to rapid changes in mean shear rates? Thanks, Casey
Hi Meangreen,
Thanks a lot for asking this question. I was planning to put this very question in the forum this morning since it was asked to me in an interview and I had no clue about it.

I understand that an explanation to this must be given in one of those turbulence modelling books, but if some of the enlightened users can tell the answer to this question it will be great.
Raashid

 April 21, 2011, 04:53 #3 Senior Member   Raashid Baig Join Date: Mar 2010 Location: Bangalore, India Posts: 135 Rep Power: 16 Can some of the enlightened users give answer to this question ? It will be very helpful. Thanks in advance, Raashid

 April 21, 2011, 22:55 #4 Senior Member   mohammad Join Date: Dec 2010 Location: UK Posts: 245 Rep Power: 16 Hi Casey, Hi Raashid, From application point of view you can find short but good info in the following references...I remember once I tried to understand the mathematical formulation and I did it somewhat ..but it was hard....I used some online notes and articles . [About the Raashid I'm sure that he knows all of the following materials....much better than me... ] 1-search for "turbulent" keyword in CFX help manual. 2- this link also covers general info about this matter http://www.cfd-online.com/Forums/mai...ga-models.html Anyway....If you find good information I also will be happy....

 April 25, 2011, 15:25 #5 Member   Casey Join Date: Jun 2009 Posts: 98 Rep Power: 17 Hey guys, I wrote a paper that gives a pretty in depth analysis and was able to learn about a lot of the differences. Basically the two-equation models are empirical and contain different terms in the dissipation equation to account for high (mean velocity) strain areas. Where can I post the description in PDF format? Thanks, Casey

April 25, 2011, 20:03
#6
Senior Member

Join Date: Dec 2010
Location: UK
Posts: 245
Rep Power: 16
Quote:
 Originally Posted by meangreen Hey guys, I wrote a paper that gives a pretty in depth analysis and was able to learn about a lot of the differences. Basically the two-equation models are empirical and contain different terms in the dissipation equation to account for high (mean velocity) strain areas. Where can I post the description in PDF format? Thanks, Casey
Hi Casey,

If the paper is online, simply give the web address...
Otherwise I think you can upload it here as an attachement to your post.
The other way is that if it's a big file, and you cant upload it here, you can upload it on such websites that are free file hosts and then share the web address ( and username and password if needed) here.

Thanks and Be Happy

April 25, 2011, 21:00
#7
Senior Member

Raashid Baig
Join Date: Mar 2010
Location: Bangalore, India
Posts: 135
Rep Power: 16
Quote:
 Originally Posted by meangreen Hey guys, I wrote a paper that gives a pretty in depth analysis and was able to learn about a lot of the differences. Basically the two-equation models are empirical and contain different terms in the dissipation equation to account for high (mean velocity) strain areas. Where can I post the description in PDF format? Thanks, Casey
Dear Casey,
As mohammad has said there are many ways of sharing the paper. If you do not want your paper to be that open you can mail your paper to selected users. Please do it as it will be very helpful in our understanding of CFD.
Raashid

 April 27, 2011, 12:30 paper #8 New Member   behzad Join Date: Aug 2010 Posts: 4 Rep Power: 15 I have read the book by Wilcox who popularized the model, "Turbulence modelling in CFD". In his book, all over, he shows how much k-omega model results are superior to the k-epsilon model in many different types of flows(near wall flows, seperated flows, ...).There is a chapter for compressible flows if you want to read. But even in that book I don't remember a term by term explanations of the differences of the modeled omega & epsilon equations. If MEANGREEN has wrote a paper about it I would very much, like to read that.

May 4, 2011, 15:54
#9
Member

Casey
Join Date: Jun 2009
Posts: 98
Rep Power: 17
Hi everyone,

I am going to try and post the latex of the Background and theory section of my paper. I may publish some of the data. Please let me know if you want to cite any part of the paper. We can work out a "private communcation" or I can tell you the citation if I got the info from somewhere else.
Attached Files
 LaMarcheBackgroundAndTheory.pdf (92.1 KB, 370 views) LaMarcheBackgroundAndTheory2.pdf (58.0 KB, 225 views)

 May 4, 2011, 15:55 #10 Member   Casey Join Date: Jun 2009 Posts: 98 Rep Power: 17 I also want to point out that the Wilcox book is very good. After doing this project I would say that the k-omega model is better than the standard k-epsilon model, but that the RNG k-epsilon model is better than the Wilcox k-omega model.

 May 5, 2011, 13:29 #11 New Member   Join Date: Oct 2010 Posts: 7 Rep Power: 15 http://personalpages.manchester.ac.u...hydr/index.htm Try this... there are a couple of lectures on turbulence modeling.

 July 7, 2012, 16:40 #12 Member   Casey Join Date: Jun 2009 Posts: 98 Rep Power: 17 If anyone uses the information that I wrote in that paper, please cite me. Thanks, Casey

July 9, 2012, 05:25
#13
Super Moderator

Sijal
Join Date: Mar 2009
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Quote:
 Originally Posted by meangreen Hi all, I am confused about the difference between the k-epsilon and k-omega turbulence models. If [math]\omega[\math] = k/, then how are the two equations so different? Can someone explain the models for me term by term and show why the k-[math]\omega[\math] model is better suited for compressible flows and modeling shear flow spreading. Why is the k-[math]\omega[\math] model less sensitive to rapid changes in mean shear rates? Thanks, Casey
Can some one clearly explain this?

It is also strange that omega is inverse to epsilon (rate/scale of turbulence) and which means they are directly related. Still there is huge difference in results for the practice flows with both models.

 July 9, 2012, 13:43 #14 Member   Casey Join Date: Jun 2009 Posts: 98 Rep Power: 17 The simplist explination is this: both the k-omega and k-epsilon models have many emperical terms. The k-omega model has terms that have been updated over the years. The k-omega model also has a different form of the turbulent viscosity equation. Page 5 of LaMarcheBackgroundandTheory.pdf above explains how the terms in the k-omega model make it more robust model. Casey